Optimize the Number of Amavis Processes
How fast Amavis can process email is directly tied to Max servers option under Warden -> Settings -> Filter Settings. More information:
Enable a Local Caching DNS Resolver
You can speed up DNS queries used by Amavis by enabling a local caching DNS resolver. More information:
Enable the Redis Storage Engine
For high volume servers we recommend switching from the default MySQL storage back end used for anti-spam bayes learning to the Redis back end. More information:
Enable the plugin
This plugin will compile the rules into native-code in order to provide significant speedups in rule evaluation. More information:
Exclude Specific Senders
If you have certain senders that are sending out large amounts of email (newsletters) then you can exclude them from anti-spam and anti-virus scanning. More information:
Disable the TxRep plugin (optional)
If your server processes a lot of email and your MySQL/MariaDB is overloaded you may want to disable the TxRep plugin as it will add about 10 additional DB queries per email and will slow down the the anti-spam learning (Amavis will still perform well with it disabled). You can disable it under Warden -> Settings -> Plugin Settings -> uncheck the TxRep checkbox and press the update button to apply the changes.
Optimizing MySQL is outside the scope of this article but we recommend that advanced users use the MySQL tuner script from the link below.
To download and run the MySQLTuner script (MySQL should have not been restarted in a few days otherwise you will not get accurate results):
wget https://raw.githubusercontent.com/major/MySQLTuner-perl/master/mysqltuner.pl -O mysqltuner.pl wget https://raw.githubusercontent.com/major/MySQLTuner-perl/master/basic_passwords.txt -O basic_passwords.txt wget https://raw.githubusercontent.com/major/MySQLTuner-perl/master/vulnerabilities.csv -O vulnerabilities.csv chmod +x mysqltuner.pl ./mysqltuner.pl
Viewing the Processing Time
To view how long a message took to be processed by Amavis go to Warden -> Logs -> Message log and click on the plus icon next to an entry. The processing time will be listed at the end on the message (in milliseconds).